Full Stack Engineer (Frontend)

Aurora surfaces a small number of roles each month — ones we'd take ourselves.Frontend / Fullstack Engineer$180K – $280K · significant equity · 22-person team · ex unicorn foundersThe companyEvery sales team has the same problem: nobody actually knows what's happening in the pipeline until it's too late.Follow-ups slip. CRM is wrong. Reps walk into calls blind. Deals die in the gap between what was discussed and what was done.We're the AI layer that closes that gap — a platform that listens across every surface where selling happens (email, calls, Slack, CRM) and acts: logging notes, auto-filling Salesforce fields, drafting follow-ups, flagging when a buyer's gone quiet, and briefing reps on deal status before they walk in.We integrate out of the box with Salesforce, HubSpot, Gmail, Slack, Zoom, and Gong.We're 22 people, well-funded by top-tier VCs, and our customers include unicorns deploying this across their entire revenue org.Founded in 2024. Growing fast.The foundersThe company is founded by repeat enterprise/product builders. The CEO co-founded Glean, now valued at $7B, after 12 years at Google working on Search Quality and Maps, and later leading engineering at Uber’s new mobility group. The co-founder founded Apple’s Cloud ML org, previously built and exited a company to Twitter, and later led Growth Engineering at Patreon as creator payouts scaled from ~$3B to $5B+.Together, they’ve shipped core products at Google, Apple, Uber, Twitter, Patreon, and Glean — and they know what it takes to build AI tools that enterprise teams actually use.The roleWe're looking for a Frontend-leaning Fullstack Engineer (3–5 YOE) to own the product surface of a platform that makes selling faster and more visible for some of the world's best revenue teams.~70% frontend, ~30% backend. The work skews heavily toward net new: new AI-powered interactions, new views, new surfaces. You will not be in a ticket queue.You'll work directly with the founders. There's no PM layer between you and product decisions.Compensation💰 $180K – $280K (higher end for exceptional ICs)📈 Significant equity — early-stage, meaningful ownership📦 Relocation support available📍 Palo Alto — hybrid, 3 days in office🛂 Visa transfers only (E3, H1B, TN). No new H1B sponsorship.What you'll buildCommitment capture and risk surfaces — the product layer that ingests a sales call, extracts every commitment and risk, and surfaces them to a rep before they forget. The problem sounds simple; the signal-to-noise problem at enterprise scale is not.AI-native CRM sync — a system that reads a conversation and fills the required Salesforce fields correctly, on its own, every time.Meeting prep interfaces — the brief a rep reads at 8:58am before a 9:00am call: deal history, open commitments, known risks, suggested talking points. Time-to-load and accuracy both matter acutely.The AI interaction layer — working directly on LLM-powered features: prompt design, output validation, graceful failure modes, and the UX that makes AI outputs feel trustworthy rather than magic.The data is messy, real-time, and mission-critical to the people using it.StackCore: React, TypeScript, Next.jsIntegrations: Salesforce, HubSpot, Gmail, Slack, Zoom, Gong — the full enterprise sales stackAI: LLM API integrations across major providers; you'll work directly on the model interaction layerInfra: AWS, modern CI/CDNice to have: Python (data pipelines), monorepo experience, prior work with LLM output validation or structured extractionWho we're looking forYou've owned a product surface end-to-end — not tickets, not features someone else designedYou care as much about whether the UX is right as whether the code is cleanYou've shipped features powered by LLM APIs and you know where they fail in productionYou've debugged a real data problem and came out with a new mental model, not just a fixYou think about the person on the other end of the interface — what they're trying to do at 8:58am, what they can't afford to get wrongYou have opinions about component architecture, data fetching, and state management — and you can defend themStrong CS fundamentals or equivalent practical depth. We care about how you think, not where you studied.Why nowSales teams are drowning in coordination overhead. The tools that exist either remind you what to do (useless if you already knew) or require you to fill in the fields yourself (same problem, different interface).The founders lived this scaling Glean and Patreon. They know the problem, they know the buyer, and they know what it takes to build AI tooling that earns trust inside a revenue org — not just a demo.The window to define this category is open right now. The engineers who join in the next few months will shape what it becomes.This is not for you ifYou want full remote or async-first workYou prefer well-defined tickets and clear swim lanesYou avoid product contextYou want a big team and established process before you shipYou're skeptical about AI as a category-defining shift in enterprise softwareInterview processFounder screen (20–30 min) — we want to understand what you've built, how you made decisions, and whyTechnical conversation (45 min) — architecture and systems thinking; no LeetCode, real problems from our stackTechnical assessment (60 min) — a problem from our actual codebasePaid work session — a short, compensated engagement so both sides can see the fit clearly before an offerWe move fast. Typically 2–3 weeks end to end.We strongly encourage applicants from underrepresented backgrounds to apply.